With the development of science and technologies, the head-mounted display has been gradually civilianized and provides high-quality visual experiences to more and more people. When an output video is displayed on the head-mounted display, light emitted from the display screen is incident into the user's eyes through a lens group, and the image viewed by the user is deformed relative to the image displayed on the display screen since an optical distortion is generated during the light propagation. In order that the user can view a normal image, a distortion processing of video data needs to be performed in advance, and the distortion-processed video data is displayed on the display screen. The distortion can offset the optical distortion generated during the light propagation in the lens group, so that the user can view a normal 2D or 3D effect.
In the prior art, a distortion processing of video data performed by the head-mounted display is through distortion processing of each image frame in the video, and specifically, distortion processing is performed on image data using FPGA in a frame unit. But due to restrictions of the FPGA storage space, the operational capability and the transmission bandwidth, a large amount of data needs to be stored for processing a singular image frame, thus the processing in a frame unit is impracticable. Meanwhile, the processing of a large amount data leads to low bandwidth utilization ratio of the distortion processing and poor distortion processing efficiency, so the video playing cannot be ensured to be smooth and the user's viewing requirement cannot be satisfied.